Meta-analyses as a privileged information source for informing teachers' practice?

Abstract. Inspired by evidence-based medicine, many researchers in the field of learning and instruction assume that meta-analyses are the best scientific information source to inform teachers' practice. This position is evaluated critically. For this purpose, I first clarify my fundamental assumptions about employing evidence for educational practice, as they form the basis for later argumentations (e.g., scientific evidence is just one of several important information sources for teachers). Then, the numerous disadvantages of meta-analyses as an information source for teachers are outlined (e.g., piecemeal information, partly inconsistent information) and, on this basis, I argue that they should not be considered a privileged source. Theories (including instructional models) provide some key advantages (e.g., coherent information) so that they should be seen as a prime information source. Nevertheless, theories also have some disadvantages so that teachers might be best advised to rely on multiple sources, and integrate them when trying to improve their practice. Finally, potential objections to theories as privileged information source are discussed.